27c2923 · typeable/bloodhound

git.iko.soy/typeable/bloodhound

Merge pull request #94 from Soostone/optimize

Implement index optimization

commit27c29236d800a45459d8f8dbec3a1999c75544e3 authorChris Allen <cma@bitemyapp.com> dateWed 3 Feb 2016, 00:05 (11 years ago) parentbfb4adf 1889edd

3 files changed+84−2

compare with master

unifiedsplit

src/Database/Bloodhound/Client.hs+47−2 src/Database/Bloodhound/Types.hs+31 tests/tests.hs+6

src/Database/Bloodhound/Client.hs+47−2

@@ -1,5 +1,6 @@

11{-# LANGUAGE OverloadedStrings #-}

22{-# LANGUAGE RecordWildCards #-}

3+{-# LANGUAGE TupleSections #-}

45-------------------------------------------------------------------------------

56-- |

@@ -26,6 +27,7 @@ module Database.Bloodhound.Client

2627 , deleteIndex

2728 , updateIndexSettings

2829 , getIndexSettings

30+ , optimizeIndex

2931 , indexExists

3032 , openIndex

3133 , closeIndex

@@ -81,7 +83,7 @@ import qualified Data.HashMap.Strict as HM

8183import Data.Ix

8284import qualified Data.List as LS (filter, foldl')

8385import Data.List.NonEmpty (NonEmpty (..))

84−import Data.Maybe (fromMaybe, isJust)

86+import Data.Maybe (catMaybes, fromMaybe, isJust)

8587import Data.Monoid

8688import Data.Text (Text)

8789import qualified Data.Text as T

@@ -302,6 +304,48 @@ getIndexSettings (IndexName indexName) = do

302304 where url = joinPath [indexName, "_settings"]

303305

304306

307+-- | 'optimizeIndex' will optimize a single index, list of indexes or

308+-- all indexes. Note that this call will block until finishing but

309+-- will continue even if the request times out. Concurrent requests to

310+-- optimize an index while another is performing will block until the

311+-- previous one finishes. For more information see

312+-- <https://www.elastic.co/guide/en/elasticsearch/reference/1.7/indices-optimize.html>. Nothing

313+-- worthwhile comes back in the reply body, so matching on the status

314+-- should suffice.

315+--

316+-- 'optimizeIndex' with a maxNumSegments of 1 and onlyExpungeDeletes

317+-- to True is the main way to release disk space back to the OS being

318+-- held by deleted documents.

319+--

320+-- Note that this API was deprecated in ElasticSearch 2.1 for the

321+-- almost completely identical forcemerge API. Adding support to that

322+-- API would be trivial but due to the significant breaking changes,

323+-- this library cannot currently be used with >= 2.0, so that feature was omitted.

324+--

325+-- >>> let ixn = IndexName "unoptimizedindex"

326+-- >>> _ <- runBH' $ deleteIndex ixn >> createIndex defaultIndexSettings ixn

327+-- >>> response <- runBH' $ optimizeIndex (IndexList (ixn :| [])) (defaultIndexOptimizationSettings { maxNumSegments = Just 1, onlyExpungeDeletes = True })

328+-- >>> respIsTwoHunna response

329+-- True

330+optimizeIndex :: MonadBH m => IndexSelection -> IndexOptimizationSettings -> m Reply

331+optimizeIndex ixs IndexOptimizationSettings {..} =

332+ bindM2 post url (return body)

333+ where url = addQuery params <$> joinPath [indexName, "_optimize"]

334+ params = catMaybes [ ("max_num_segments",) . Just . showText <$> maxNumSegments

335+ , Just ("only_expunge_deletes", Just (boolQP onlyExpungeDeletes))

336+ , Just ("flush", Just (boolQP flushAfterOptimize))

337+ ]

338+ indexName = indexSelectionName ixs

339+ boolQP True = "true"

340+ boolQP False = "false"

341+ body = Nothing

342+

343+

344+-------------------------------------------------------------------------------

345+indexSelectionName :: IndexSelection -> Text

346+indexSelectionName (IndexList names) = T.intercalate "," [n | IndexName n <- toList names]

347+indexSelectionName AllIndexes = "_all"

348+

305349deepMerge :: [Object] -> Object

306350deepMerge = LS.foldl' go mempty

307351 where go acc = LS.foldl' go' acc . HM.toList

@@ -395,6 +439,7 @@ closeIndex = openOrCloseIndexes CloseIndex

395439-- >>> let aliasName = IndexName "an-alias"

396440-- >>> let iAlias = IndexAlias src (IndexAliasName aliasName)

397441-- >>> let aliasCreate = IndexAliasCreate Nothing Nothing

442+-- >>> _ <- runBH' $ deleteIndex src

398443-- >>> respIsTwoHunna <$> runBH' (createIndex defaultIndexSettings src)

399444-- True

400445-- >>> runBH' $ indexExists src

@@ -484,7 +529,7 @@ versionCtlParams cfg =

484529 ExternalGTE (ExternalDocVersion v) -> versionParams v "external_gte"

485530 ForceVersion (ExternalDocVersion v) -> versionParams v "force"

486531 where

487− vt = T.pack . show . docVersionNumber

532+ vt = showText . docVersionNumber

488533 versionParams v t = [ ("version", Just $ vt v)

489534 , ("version_type", Just t)

490535 ]

src/Database/Bloodhound/Types.hs+31

@@ -141,6 +141,9 @@ module Database.Bloodhound.Types

141141 , FieldName(..)

142142 , Script(..)

143143 , IndexName(..)

144+ , IndexSelection(..)

145+ , IndexOptimizationSettings(..)

146+ , defaultIndexOptimizationSettings

144147 , TemplateName(..)

145148 , TemplatePattern(..)

146149 , MappingName(..)

@@ -384,6 +387,28 @@ data IndexSettings =

384387defaultIndexSettings :: IndexSettings

385388defaultIndexSettings = IndexSettings (ShardCount 3) (ReplicaCount 2)

386389

390+

391+{-| 'IndexOptimizationSettings' is used to configure index optimization. See

392+ <https://www.elastic.co/guide/en/elasticsearch/reference/1.7/indices-optimize.html>

393+ for more info.

394+-}

395+data IndexOptimizationSettings =

396+ IndexOptimizationSettings { maxNumSegments :: Maybe Int

397+ -- ^ Number of segments to optimize to. 1 will fully optimize the index. If omitted, the default behavior is to only optimize if the server deems it necessary.

398+ , onlyExpungeDeletes :: Bool

399+ -- ^ Should the optimize process only expunge segments with deletes in them? If the purpose of the optimization is to free disk space, this should be set to True.

400+ , flushAfterOptimize :: Bool

401+ -- ^ Should a flush be performed after the optimize.

402+ } deriving (Eq, Show, Generic, Typeable)

403+

404+

405+{-| 'defaultIndexOptimizationSettings' implements the default settings that

406+ ElasticSearch uses for index optimization. 'maxNumSegments' is Nothing,

407+ 'onlyExpungeDeletes' is False, and flushAfterOptimize is True.

408+-}

409+defaultIndexOptimizationSettings :: IndexOptimizationSettings

410+defaultIndexOptimizationSettings = IndexOptimizationSettings Nothing False True

411+

387412{-| 'UpdatableIndexSetting' are settings which may be updated after an index is created.

388413

389414 <https://www.elastic.co/guide/en/elasticsearch/reference/current/indices-update-settings.html>

@@ -770,6 +795,12 @@ newtype ReplicaCount = ReplicaCount Int deriving (Eq, Show, Generic, ToJSON, Typ

770795-}

771796newtype IndexName = IndexName Text deriving (Eq, Generic, Show, ToJSON, FromJSON, Typeable)

772797

798+{-| 'IndexSelection' is used for APIs which take a single index, a list of

799+ indexes, or the special @_all@ index.

800+-}

801+data IndexSelection = IndexList (NonEmpty IndexName)

802+ | AllIndexes deriving (Eq, Generic, Show, Typeable)

803+

773804{-| 'TemplateName' is used to describe which template to query/create/delete

774805-}

775806newtype TemplateName = TemplateName Text deriving (Eq, Show, Generic, ToJSON, FromJSON, Typeable)

tests/tests.hs+6

@@ -1446,6 +1446,12 @@ main = hspec $ do

14461446 (IndexSettings (ShardCount 1) (ReplicaCount 0))

14471447 (NE.toList updates))

14481448

1449+ describe "Index Optimization" $ do

1450+ it "returns a successful response upon completion" $ withTestEnv $ do

1451+ _ <- createExampleIndex

1452+ resp <- optimizeIndex (IndexList (testIndex :| [])) defaultIndexOptimizationSettings

1453+ liftIO $ validateStatus resp 200

1454+

14491455 describe "JSON instances" $ do

14501456 propJSON (Proxy :: Proxy Version)

14511457 propJSON (Proxy :: Proxy IndexName)